In the fields of marketing and security, understanding the gender, age, address, and other attributes of individuals captured by security cameras is crucial. However, camera images alone provide only appearance information, making it difficult to accurately determine personal attributes. While biometric authentication, such as facial recognition or iris scans, has become mainstream for identifying individuals on camera, this approach requires prior biometric registration, which is challenging when dealing with large, unspecified populations. To address this issue, we focused on smartphones, widely carried by individuals, as a potential identification tool. Smartphones already store substantial biometric and personal information, such as facial and fingerprint data, leading us to consider them for identity verification. In this paper, we propose a method for personal authentication that does not require prior biometric registration. Our approach pairs each individual captured by a 360-degree camera with their respective smartphone. The pairing is achieved by having the individual shake the smartphone, held in their right hand, in front of the camera, using the smartphone’s accelerometer data. We believe this method has applications in advanced behavior analysis, employee management, and suspicious person detection, as it effectively combines vision-based human activity recognition (HAR) with sensor-based HAR.